Control rod homogenization in heterogeneous sodium-cooled fast reactors

The sodium-cooled fast reactor is one of the candidates for a sustainable nuclear reactor system. In particular, the French ASTRID project employs an axially heterogeneous design, proposed in the so-called CFV (low sodium effect) core, to enhance the inherent safety features of the reactor. This thesis focuses on the accurate modeling of the control rods, through the homogenization method. The control rods in a sodium-cooled fast reactor are used for reactivity compensation during the cycle, power shaping, and to shutdown the reactor. In previous control rod homogenization procedures, only a radial description of the geometry was implemented, hence the axially heterogeneous features of the CFV core could not be taken into account. This thesis investigates the different axial variations the control rod experiences in a CFV core, to determine the impact that these axial environments have on the control rod modeling. The methodology used in this work is based on previous homogenization procedures, the so-called equivalence procedure. The procedure was newly implemented in the PARIS code system in order to be able to use 3D geometries, and thereby be take axial effects into account. The thesis is divided into three parts. The first part investigates the impact of different neutron spectra on the homogeneous control-rod cross sections. The second part investigates the cases where the traditional radial control-rod homogenization procedure is no longer applicable in the CFV core, which was found to be 5-10 cm away from any material interface. In the third part, based on the results from the second part, a 3D model of the control rod is used to calculate homogenized control-rod cross sections. In a full core model, a study is made to investigate the impact these axial effects have on control rod-related core parameters, such as the control rod worth, the capture rates in the control rod, and the power in the adjacent fuel assemblies. All results were compared to a Monte Carlo-based model which served as the reference. It was demonstrated that the treatment of the radial environment, surrounding the control rod, has a small impact on the core parameters, and the traditional radial homogenization procedure yields reliable results. For axial interfaces within the control rod itself, the traditional radial homogenization procedure could no longer capture the large environmental impact, hence 3D modeling is recommended. (author)